University of Groningen
Quantitative cardiac dual source CT; from morphology to function
Assen, van, Marly
DOI:
10.33612/diss.93012859
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Assen, van, M. (2019). Quantitative cardiac dual source CT; from morphology to function. Rijksuniversiteit Groningen. https://doi.org/10.33612/diss.93012859
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Feasibility of Extracellular Volume
Quantification using Dual-energy CT
Technical Report
Marly van Assen, Carlo N. De Cecco, PhD, Pooyan Sahbaee, Marwen Eid, MD, L. Parkwood Griffith, BS, Rock H. Savage, Akos Varga-Szemes, Matthijs Oudkerk, Rozemarijn Vliegenthart, U. Joseph Schoepf
Published JCCT 2018
ABSTRACT
Objective: To assess the feasibility of dual energy CT (DECT) to derive myocardial extracellular volume (ECV) and detect myocardial ECV differences without a non-contrast acquisition, compared to single energy CT (SECT).
Methods: Subjects (n=35) with focal fibrosis (n=17), diffuse fibrosis (n=10), and controls (n=9) underwent non-contrast and delayed acquisitions to calculate SECT-ECV. DECT-ECV was calculated using the delayed acquisition and the derived virtual non-contrast images. In the control and diffuse fibrotic groups, the entire myocardium of the left ventricle was used to calculate ECV. Two ROIs were placed in the focal fibrotic group, one in normal and one in fibrotic myocardium.
Results: Median ECV was 33.4% (IQR, 30.1-37.4) using SECT and 34.9% (IQR, 31.2-39.2) using DECT (p=0.401). For both techniques, focal and diffuse fibrosis had significantly higher ECV values (all p<0.021) than normal myocardium. There was no systematic bias between DECT and SECT (p=0.348). SECT had a higher radiation dose (1.1mSv difference) than DECT (p<0.001).
Conclusion: ECV can be measured using a DECT approach with only a delayed acquisition. The DECT approach provides similar results at a lower radiation dose compared to SECT.
INTRODUCTION
Recent studies have shown that extracellular volume (ECV) fraction can be derived from non-contrast and delayed-phase CT with high histology and MRI-derived ECV correlation.(1–4) Compared to MRI, CT offers a faster acquisition, the potential to scan patients with metal implants, and the assessment of coronary anatomy. However, single-energy CT (SECT) is less sensitive to contrast differences than MRI, has a poor contrast-to-noise ratio, and suffers from artifacts (mainly caused by noise), especially in the delayed enhancement acquisition. The dual-energy CT (DECT) technique uses two different kV levels, allowing for tissue characterization(5,6) and potential measurement of ECV for the detection of diffuse myocardial fibrosis in cardiomyopathies.(7–9) However, no studies have shown the possibility of detecting focal fibrosis, and no direct comparison with SECT has been performed.
Thus, the purpose of this study was to assess the feasibility of DECT to derive ECV and detect differences between myocardial ECV, in order to compare the results with SECT.
METHODS
PopulationWe analyzed data of patients (n=42) who underwent cardiac DECT between 2008-2017 for the evaluation of suspected CAD or known cardiomyopathy, who also underwent a non-contrast acquisition in addition to a contrast-enhanced rest and delayed DECT acquisition. Written informed consent had been obtained from all patients.
Based on their clinical history and the image findings, patients were assigned to three groups: 1) controls; 2) focal fibrotic (chronic myocardial infarct); or 3) diffuse fibrotic (cardiomyopathies). The focal fibrotic and control groups were selected based on MRI. Presence and location of infarction was defined based on late gadolinium enhancement (LGE) MRI. Recent MRIs were not available in the diffuse group because of the prevalence of intracardiac devices. Patients in the diffuse fibrotic group all had a previous MRI-based diagnosis of a cardiomyopathy associated with diffuse ECV increase.
Image Acquisition
All examinations were performed with a second- or third-generation dual-source CT system (Definition Flash or Force; Siemens Healthineers, Forchheim, Germany). The DECT protocol included a non-contrast acquisition (prospective ECG-triggering, tube voltage 120kV, tube current 75-80mAs with automated tube current modulation [CARE
chapter13: feasibilityofextracellularvolumequantificationusingdual-energyct
Dose 4D, Siemens], slice thickness 3mm and increment 1.5mm), a DECT contrast-enhanced acquisition at rest (prospective ECG gating, tube voltage and current, 100/140kVp with 165 reference mA/rotation [second-generation] and 90/150kVp with Tin filtration and 90mA/rotation [third-generation], gantry rotation time 280ms/250ms [second/third-generation], heart rate dependent pitch 0.2–0.43, and 1.5mm section thickness with 1mm overlap reconstructed in 60-75% diastole), and a DECT delayed acquisition 7 minutes after contrast injection (same DECT protocol). Contrast (70mL iopromide [370mgI/mL; Ultravist, Bayer, Wayne, NJ]) was administered at a flow rate of 5.0mL/s using a dual-syringe injector (Stellant D, Medrad, Indianola, PA) and automated bolus triggering.
LGE MRI protocol
For the LGE MRI acquisition, a standard protocol was used 12 minutes after the administration of contrast agent (0.1mmol/kg gadobenate-dimeglumine, MultiHance, Bracco, Princeton, NJ). MRI scans were evaluated by two board-certified physicians. DECT Post-processing
Three different image sets were derived from the DECT data on a 3D workstation (syngo.via VB10B, Siemens), see Figure 1. The low- and high-kVp datasets were used to generate the iodine maps and the virtual non-contrast images. The standard linearly blended image set was created by combining 60% high-kVp and 40% low-kVp data, representing a standard 120kVp SECT acquisition. The fourth image set is the real SECT non-contrast image.
Figure 1: Schematic overview of image reconstruction. Black arrows indicate datasets used for
Image Analysis
For SECT, ECV was calculated using the true non-contrast images and merged virtual delayed images. For DECT, the virtual non-contrast and delayed scans were used. ECV was calculated in all groups for both SECT and DECT. In the control and diffuse fibrotic groups, a region of interest (ROI) encompassing the entire left ventricle was used to calculate ECV of normal and fibrotic myocardium. In the focal fibrotic group, two ROIs (at least 1cm2 each) were placed, one in the normal myocardium and one
in the fibrotic myocardium. The slice with the largest infarct area was chosen for measurements. The areas of infarction were matched with LGE MRI by an experienced observer.
The contrast-enhanced acquisitions and LGE MRI were used to ensure correct placement of the ROIs in the non-contrast and delayed datasets. Finally, a ROI was placed in the left ventricle of each dataset to measure the blood pool. For each of these ROIs, the HU values were recorded for the delayed DECT and the true and virtual non-contrast and delayed SECT.
ECV was calculated using the following equation:
Image Analysis
For SECT, ECV was calculated using the true non-contrast images and merged virtual
delayed images. For DECT, the virtual non-contrast and delayed scans were used.
ECV was calculated in all groups for both SECT and DECT. In the control and diffuse
fibrotic groups, a region of interest (ROI) encompassing the entire left ventricle was
used to calculate ECV of normal and fibrotic myocardium. In the focal fibrotic group,
two ROIs (at least 1cm
2each) were placed, one in the normal myocardium and one in
the fibrotic myocardium. The slice with the largest infarct area was chosen for
measurements. The areas of infarction were matched with LGE MRI by an experienced
observer.
The contrast-enhanced acquisitions and LGE MRI were used to ensure correct
placement of the ROIs in the non-contrast and delayed datasets. Finally, a ROI was
placed in the left ventricle of each dataset to measure the blood pool. For each of these
ROIs, the HU values were recorded for the delayed DECT and the true and virtual
non-contrast and delayed SECT.
ECV was calculated using the following equation:
ECV
CT= (1 − hematocrit) ×
ΔHU
ΔHU
myocardium bloodpoolwhere ΔHU
myocardiumis the difference between HU values of the myocardium at
(virtual) non-contrast images and delayed images, and ΔHU
bloodpoolis the difference
between HU values of the blood pool at (virtual) non-contrast images and delayed
images. The most recent hematocrit value was recorded from clinical records.
319
where ΔHUmyocardium is the difference between HU values of the myocardium at (virtual) non-contrast images and delayed images, and ΔHUbloodpool is the difference between HU values of the blood pool at (virtual) non-contrast images and delayed images. The most recent hematocrit value was recorded from clinical records.
Statistical analysis
Normal distribution was assessed using the Shapiro-Wilk test. Continuous variables were expressed as mean±SD; and non-parametric variables were expressed as median with inter-quartile ranges. A Chi2 test for multiple groups was used to compare
demographics. A Wilcoxon signed rank test was used to assess differences between DECT-ECV and SECT-ECV. An independent Mann-Whitney U test examined differences between ECV measurements in normal myocardium and infarcted myocardium. Bland–Altman analysis was used to demonstrate agreement between the SECT and DECT approach. A statistical package (SPSS, version 24) was used for all data analysis. A p-value <0.05 was considered statistically significant.
chapter13: feasibilityofextracellularvolumequantificationusingdual-energyct
RESULTS
Patient population
Seven patients were excluded from the study due to unavailable non-contrast, rest or delayed DECT acquisitions (n=3), inadequate image quality (n=2) from motion artefacts (1) and high noise levels (1), or small (<1cm2) LGE area on MRI (n=2). No patients were
excluded because of metal wire artifacts. Thus, a total of 35 patients were included and 104 ECV measurements were analyzed. There were no significant differences in the patient demographics (Table 1) among the three groups, except for smoking (p=0.046). Hematocrit values showed a trend to be lower in the diseased groups than the control group (p=0.50). The mean radiation doses for SECT and were 13.2±6.3mSv and 12.1±5.9mSv, respectively (p<0.001).
Table 1. Patient Demographics
Control
(n=8) Focal fibrotic(n=17) Diffuse fibrotic(n=10)
Age (year) 66 (57-72) 66 (49-71) 65 (45-70) Gender (male) 5 (63) 16 (95) 7 (70) BMI (kg/m2) 31.1 (24.8-37.1) 26.8 (22.6-30.5) 29.2 (22.1-34.4) Hematocrit (%) 43 (39-46) 41 (39-42) 39 (35-43) Diabetes 4 (50) 3 (18) 4 (40) Hypertension 6 (75) 13 (77) 7 (70) Smoking 4 (50) 6 (35) 0 * Hyperlipidemia 5 (63) 13 (77) 7 (70) Metallic wires 0 (0) 0 (0) 9 (90)
Data reported as medians with IQR or as frequencies and percentages. (*) indicates significant differences compared to the control group.
ECV measurements
The median ECV of normal myocardium measured with SECT and DECT were 33.4% (IQR, 30.1-37.4) and 34.9% (IQR, 31.2-39.2), respectively (p=0.401) (Table 2). ECV in the normal myocardium of the control group was slightly, but not significantly, higher than that of the normal myocardium in the focal fibrotic group for both SECT and DECT (Table 2).
In the focal fibrotic group, no significant difference was found between SECT-ECV and DECT-ECV values (p=0.943) (Table 2). Infarcted myocardium had significantly higher ECV values (p<0.001) than normal myocardium in both SECT and DECT.
The diffuse fibrotic group included patients with sarcoidosis (n=2), ischemic cardiomyopathy (n=3), arrhythmogenic cardiomyopathy (n=2), and non-ischemic dilated cardiomyopathy (n=3). There was no significant difference between SECT-ECV and DECT-ECV values (p=0.575) (Table 2). Diffuse fibrotic SECT-ECV and DECT-ECV were significantly higher than ECV in normal myocardium (p<0.001 and p=0.021, respectively), but significantly lower than focal fibrotic myocardium (both p<0.001). Bland-Altman analysis between SECT-ECV and DECT-ECV showed no systematic bias between measurements (p=0.348) with a limits of agreement of ±9.4% (Figure 2).
Table 2. Overview of ECV measurements for normal, focal fibrosis and diffuse fibrosis
Normal myocardium Diseased myocardium
SECT DECT P SECT DECT p p
Diseased vs control Control 35.4% (32.4-36.3) 35.1%(33.4-40.2) 0.401 - -Focal fibrotic 32.0%(28.9-36.3) 33.6%(29.3-39.2) 0.467 48.4%*(46.4-50.7) 49.4%*(46.3-50.6) 0.943 <0.001 (SECT) <0.001 (DECT) Diffuse Fibrotic - - 40.7%*(38.0-43.5) 39.4%*(37.0-41.6) 0.575 <0.001 (SECT) 0.021 (DECT)
Data reported as percentages. (*) indicates significantly different ECV compared to normal myocardium
chapter13: feasibilityofextracellularvolumequantificationusingdual-energyct
Figure 2: Bland-Altman analysis showed no systematic bias between SECT and DECT.
DISCUSSION
For this study we evaluated the feasibility of DECT to calculate ECV using virtual non-contrast images compared to SECT-based true non-contrast acquisitions. There were no systematic differences between SECT and DECT. Both approaches were able to discriminate diseased and healthy myocardium, where focal fibrosis had significantly higher ECV compared to normal myocardium and diffuse fibrosis. This preliminary study shows the feasibility of using DECT for ECV measurement with a decreased number of acquisitions, i.e. eliminating the need for a true non-contrast scan, thus reducing radiation dose.
Hong et al. recently reported on the use of DECT to characterize changes in myocardial tissue within an animal model of doxorubicin-induced cardiomyopathy. Their result showed that the ECV calculated from DECT has excellent agreement with histological and CMR determined ECV. However, this study used the iodine map with corresponding overlay HU values, relying on a single HU value instead of using the difference in HU values between virtual non-contrast and delayed acquisitions (7). The values found in our study, both with SECT and DECT, are higher than ECV measurements (mean ECV ranging between 26.6 and 31%) previously presented in studies using single energy acquisitions (2,3). This difference could be caused by the discrepancy in the ECV equations that were used. In our study, we opted to use the same approach as studies
that utilized SECT and CMR, and also took advantage of DECT to construct a virtual non-contrast image, eliminating the need of a true non contrast acquisition. Future studies should investigate the difference between using this approach and the approach proposed by Hong et al. (7). Another reason for this difference in ECV values could be the time-delay between contrast administration and the delayed image acquisition. In previous SECT-ECV studies, the time-delay between contrast administration and the delayed acquisition was set between 10 and 25 minutes after contrast injection while 12-minutes delay were used in a previous DECT study (8). In our DECT study a delay of 7 minutes was chosen. This delay is shorter than those in previous studies, potentially resulting in a bigger difference between HU values pre and post-contrast in the myocardium, which would ultimately result in increased ECV values. However, an animal study investigating multiple delays revealed no significant changes in the ECV results between 3 and 20 minutes of time delay (7).
An overestimation of ECV is also shown in previous studies using a SECT-ECV approach compared to CMR results. Myocardial ECV measured with CMR at 1.5 T in normal controls was reported to be 21-28%; these values are significantly lower than the ECV values (with a median of 35%) assessed with CT in normal controls reported in this study (10–12). Previous reviews on ECV mapping for tissue characterization show that focal fibrosis, as in myocardial infarction, has higher ECV values than diffuse fibrosis, as seen in hypertrophic and dilated cardiomyopathies. In this study, we see the same difference using both the single energy and the dual energy approach.
Using a DECT approach to evaluate ECV does not only provide similar results to the SECT approach, it also results in a decrease in radiation dose by eliminating the need for a non-contrast scan and the use of a contrast enhanced scan to determine the myocardial contours. This will also lead to a decrease in calculation time. Using only one acquisition for the calculation of ECV greatly improves the possibility of developing a fully automatic ECV algorithm. This is due to the fact that using a DECT approach avoids the mismatch error of drawing ROIs at different positions caused by the separate acquisition of the non-contrast and delayed scan used in a SECT approach. Compared to SECT acquisitions, DECT offers the possibility to evaluate scans at different kV levels and has the potential to reduce both beam hardening and metal artefacts (13–15). Among the patients included in this study with diffuse fibrosis, almost all (90%) had pacemakers or implantable cardioverter defibrillator (ICD) wires at the time of their CT acquisition, resulting in metal artifacts. Metallic devices greatly influence the image quality of SECT datasets, causing severe beam hardening and photon starvation artifacts. Both of these artifacts are intrinsically related to the polychromatic nature of the single-energy x-ray beam used in SECT examinations. Strategies to reduce beam
chapter13: feasibilityofextracellularvolumequantificationusingdual-energyct
hardening artifacts rely on the development of specific algorithms and the use of higher tube potentials. These strategies come with several disadvantages such as an increase in radiation dose (16–18). The use of dual energy offers the unique ability to reduce beam hardening artifacts with the simple post-processing procedure of adjusting the monoenergetic level to the optimal value (19,20). Several studies on patients with metallic implants that underwent DECT showed that monoenergetic level optimization provided superior image quality and diagnostic value (16,17). The DECT approach is of particular interest in the metallic device population since this group of patients is excluded from CMR examination.
Limitations
There are some limitations to this current investigation that deserve mention. First, a limited number of patients were included, resulting in only 50 ECV measurements in normal myocardium. Furthermore, the main objective of this study was to evaluate the potential of DECT to evaluate ECV changes in patients with focal and diffuse fibrotic tissue compared to a single energy approach. Therefore, a reference standard, such as pathology or CMR ECV, was not included.
In conclusion, it is feasible to use a DECT approach to measure ECV using only a delayed acquisition. The DECT approach provides similar results compared to a SECT approach at a lower radiation dose and potential workflow advantages.
REFERENCES
1. Kurita Y, Kitagawa K, Kurobe Y, Nakamori S, Nakajima H, Dohi K, et al. Estimation of myocardial extracellular volume fraction with cardiac CT in subjects without clinical coronary artery disease: A feasibility study. J Cardiovasc Comput Tomogr [Internet]. 2016;10(3):237–41. Available from: http://dx.doi.org/10.1016/j.jcct.2016.02.001
2. Bandula S, White SK, Flett AS, Lawrence D, Pugliese F, Ashworth MT, et al. Measurement of Myocardial Extracellular Volume Fraction by Using Equilibrium Contrast-enhanced CT: Validation against Histologic Findings. Radiology [Internet]. 2013;269(2):396–403. Available from: http://pubs.rsna.org/doi/10.1148/radiol.13130130
3. Nacif MS, Kawel N, Lee JJ, Zavodni A, Sibley CT, Lima JAC. Interstitial Myocardial Fibrosis Assessed as Extracellular Volume Fraction with Low-Radiation-Dose Cardiac CT. Radiology. 2012;264(3).
4. Distefano MD. 3D left ventricular extracellular volume fraction by low-radiation dose cardiac CT: Assessment of interstitial myocardial fibrosis. J Cardiovasc Comput Tomogr. 2015;7(5):213– 23.
5. Ruzsics B, Lee H, Powers ER, Flohr TG, Costello P, Schoepf UJ. Myocardial ischemia diagnosed by dual-energy computed tomography: Correlation with single-photon emission computed tomography. Circulation. 2008;117(9):1244–5.
6. Nakahara, T., Toyama, T., Jinzaki, M., Seki, R., Saito, Y., Higuchi, T., Yamada, M., Arai, M., Tsushima, Y., Kuribayashi, S., Kurabayashi M. Quantitative Analysis of Iodine Image of Dual-energy Computed Tomography at Rest: Comparison With 99mTc-Tetrofosmin Stress-rest Single-photon Emission Computed Tomography Myocardial Perfusion Imaging as the Reference Standard. J Thorac Imaging. 2018;33(2):97–104.
7. Hong YJ, Kim TK, Hong D, Park CH, Yoo SJ, Wickum ME, et al. Myocardial Characterization Using Dual-Energy CT in Doxorubicin-Induced DCM: Comparison With CMR T1-Mapping and Histology in a Rabbit Model. JACC Cardiovasc Imaging. 2016;9(7):836–45.
8. Lee H, Im DJ, Youn J, Chang S, Suh YJ, Hong YJ, et al. Myocardial Extracellular Volume Fraction with Dual-Energy Equilibrium Contrast-enhanced Cardiac CT in Nonischemic Cardiomyopathy: A Prospective Comparison with Cardiac MR Imaging. Radiology [Internet]. 2016;280(1):49–57. Available from: http://pubs.rsna.org/doi/10.1148/radiol.2016151289
9. Wang R, Liu X, Schoepf UJ, van Assen M, Alimohamed I, Griffith LP, et al. Extracellular volume quantitation using dual-energy CT in patients with heart failure: Comparison with 3T cardiac MR. Int J Cardiol. 2018;
10. Hamlin SA, Henry TS, Little BP, Lerakis S, Stillman AE. Mapping the Future of Cardiac MR Imaging: Case-based Review of T1 and T2 Mapping Techniques. RadioGraphics. 2014;34:1594– 611.
11. Muscogiuri G, Suranyi P, Schoepf UJ, De Cecco CN, Secinaro A, Wichmann JL, et al. Cardiac Magnetic Resonance T1-Mapping of the Myocardium. J Thorac Imaging [Internet]. 2017;00(00):1. Available from: http://insights.ovid.com/crossref?an=00005382-900000000-99633
12. Haaf P, Garg P, Messroghli DR, Broadbent DA, Greenwood JP, Plein S. Cardiac T1 Mapping and Extracellular Volume (ECV) in clinical practice: a comprehensive review. J Cardiovasc Magn Reson. 2017;18(1):89.
13. Yu L, Leng S, McCollough CH. Dual-energy CT-based monochromatic imaging. AJR Am J Roentgenol. 2012;199(5 Suppl):9–15.
14. Johnson TRC. Dual-energy CT: general principles. AJR Am J Roentgenol. 2012;199(5 Suppl):3–8. 15. Vliegenthart R, Pelgrim GJ, Ebersberger U, Rowe GW, Oudkerk M, Schoepf UJ. Dual-energy CT
of the heart. AJR Am J Roentgenol. 2012;199(5 Suppl):54–63.
chapter13: feasibilityofextracellularvolumequantificationusingdual-energyct
16. Bamberg F, Dierks A, Nikolaou K, Reiser MF, Becker CR, Johnson TRC. Metal artifact reduction by dual energy computed tomography using monoenergetic extrapolation. Eur Radiol. 2011;21(7):1424–9.
17. Meinel FG, Bischoff B, Zhang Q, Bamberg F, Reiser MF, Johnson TRC. Metal artifact reduction by dual-energy computed tomography using energetic extrapolation: a systematically optimized protocol. Invest Radiol [Internet]. 2012;47(7):406–14. Available from: http://www.ncbi.nlm.nih. gov/pubmed/22659595
18. Watzke O, Kalender WA. A pragmatic approach to metal artifact reduction in CT: Merging of metal artifact reduced images. Eur Radiol. 2004;14(5):849–56.
19. Tabari A, Lo Gullo R, Murugan V, Otrakji A, Digumarthy S KM. Recent Advances in Computed Tomographic Technology: Cardiopulmonary Imaging Applications. J Thorac Imaging. 2017;32(2):89–100.
20. Lenga L, Albrecht MH, Othman AE, Martin SS, Leithner D, D’Angelo T AC, Scholtz JE, De Cecco CN, Schoepf UJ, Vogl TJ WJ. Monoenergetic Dual-energy Computed Tomographic Imaging: Cardiothoracic Applications. J Thorac Imaging. 2017;32(3):151–8.
chapter13: feasibilityofextracellularvolumequantificationusingdual-energyct
